Soatto, Stefano and Perona, Pietro (1996) Reducing "Structure From Motion": a General Framework for Dynamic Vision - Part 2: Experimental Evaluation. California Institute of Technology , Pasadena, CA. (Unpublished) http://resolver.caltech.edu/CaltechCDSTR:1996.002
See Usage Policy.
See Usage Policy.
Use this Persistent URL to link to this item: http://resolver.caltech.edu/CaltechCDSTR:1996.002
A number of methods have been proposed in the literature for estimating scene-structure and ego-motion from a sequence of images using dynamical models. Although all methods may be derived from a "natural" dynamical model within a unified framework, from an engineering perspective there are a number of trade-offs that lead to different strategies depending upon the specific applications and the goals one is targeting. Which one is the winning strategy? In this paper we analyze the properties of the dynamical models that originate from each strategy under a variety of experimental conditions. For each model we assess the accuracy of the estimates, their robustness to measurement noise, sensitivity to initial conditions and visual angle, effects of the bas-relief ambiguity and occlusions, dependence upon the number of image measurements and their sampling rate.
|Item Type:||Report or Paper (Technical Report)|
|Group:||Control and Dynamical Systems Technical Reports|
|Subject Keywords:||Visual motion estimation, epipolar geometry, motion decoupling, fixation, parallax, model reduction, implicit Extended Kalman Filter|
|Usage Policy:||You are granted permission for individual, educational, research and non-commercial reproduction, distribution, display and performance of this work in any format.|
|Deposited By:||Imported from CaltechCDSTR|
|Deposited On:||20 Oct 2006|
|Last Modified:||26 Dec 2012 14:30|
Repository Staff Only: item control page